CN110736539A - gaze type spectral imaging system based on compressed sensing - Google Patents

Abstract

The invention provides a gaze-type spectral imaging system based on compressed sensing, which solves the problems of increased data transmission cost, low signal-to-noise ratio, large system volume, high cost and poor imaging quality of the existing imaging system. The system includes a first imaging mirror, a light splitting unit, a first encoding template, a light combining unit, a second encoding template, an imaging mirror group, a unit detector and a data processing unit arranged in sequence along the beam direction; the first imaging mirror is used to target imaging; the light splitting unit includes a first collimating mirror, a first dispersion element, and a second imaging mirror; the first encoding template is used to perform spectral dimension encoding on the image information imaged by the second imaging mirror; the light combining unit includes a second collimating mirror A straight mirror, a second dispersive element, and a third imaging mirror; the second encoding template is used to perform spatial micro-coding on the image information imaged by the third imaging mirror; the imaging mirror group is used to compress the encoded image to the unit detector focus On the plane, the data processing unit is used to restore the target information.

Description

A gaze-based spectral imaging system based on compressed sensing

technical field

The invention relates to spectral imaging technology, in particular to a gaze-type spectral imaging system based on compressed sensing.

Background technique

Hyper/multi-spectral imaging technology can obtain the image information and spectral information of the target, and obtain the surface structure and spectral characteristic information of the target. , chemical composition monitoring and many other fields, it has indispensable application value in military, industrial and agricultural production. With the development of spectral imaging technology, its application bands have been continuously expanded from visible light and near-infrared to mid- and long-wave infrared and terahertz bands.

The traditional single-channel spectral imaging technology has a low signal-to-noise ratio. Compared with traditional optical imaging, the spectral imager can obtain a three-dimensional data cube of the target. The amount of data will bring about data storage problems, and the bandwidth tolerance characteristics required for the signal to transmit information will increase with water. The cost of transmission increases and the hardware requirements increase, otherwise the data processing time will be seriously affected and the work efficiency will be affected. For airborne and spaceborne equipment, there will also be problems of limited information acquisition and transmission, resulting in the inability to obtain detection and reconnaissance data in real time; Large-scale detectors require higher costs; while in the terahertz and mid- and long-wave infrared bands, the technology of large-scale detectors is immature, expensive, and difficult to guarantee performance, and due to the mid-wave infrared and terahertz imaging semiconductor technology The development is relatively slow, so it is necessary to change the thinking of large area array detection. The concept of compressed sensing makes single-pixel imaging possible. Compressive Sensing (CS), also known as compressed sampling, is a new mathematical theory proposed by E. Candès, J. Romberg, T. Tao and Donoho in recent years. A linear, non-adaptive global observation is performed on the original signal to obtain a small amount of observation signal, and then the original signal is accurately reconstructed by a reconstruction algorithm. In 2006, the single-pixel camera at Rice University in the United States was also successfully introduced with the improvement of compressed sensing theory. The working principle of this device is actually the basic application of compressed sensing theory in optical imaging. Compared with the traditional imaging method of one-to-one correspondence , and obtain picture information by increasing the light intensity values corresponding to some pixel blocks with overall scene information. Refer to the research results of Rice University. In 2008, terahertz single-pixel imaging was successfully achieved. In 2011, Correia M's research group successfully developed a single-pixel camera with dynamic illumination. In 2013, B.Sun et al. successfully developed a single-pixel camera with 3D imaging function. In 2016, Yiwei Zhang et al. designed a single-pixel 3D video imaging device; in 2017, Chi H et al. used a spectrometer to achieve single-pixel spectral imaging; Zhang Z et al. proposed a modulation mode based on Fourier transform to achieve single-pixel imaging. imaging.

To sum up, the current research on single-pixel imaging mainly focuses on traditional panchromatic imaging, and some studies extend it to 3D imaging and spectral imaging, but these imaging processes all need to use scanning to obtain the three-dimensional information of the target. The scanning process requires motors, control devices and moving parts, which increases the overall volume, weight and cost of the system, while reducing system reliability and imaging quality.

SUMMARY OF THE INVENTION

In order to solve the technical problems of increased data transmission cost and low signal-to-noise ratio caused by the increase in the amount of data acquired by the existing imaging system, and the existing single-pixel imaging scanning method caused the system to be large in size, high in cost, and poor in imaging quality, the present invention provides A gaze-based spectral imaging system based on compressed sensing.

For achieving the above object, the technical scheme provided by the present invention is:

A gaze-type spectral imaging system based on compressed sensing, which is special in that it includes a first imaging mirror, a light splitting unit, a first encoding template, a light combining unit, a second encoding template, and an imaging mirror group arranged in sequence along the beam direction. , a unit detector and a data processing unit; the first imaging mirror is used to image the target at the primary image plane position; the spectroscopic unit includes a first collimating mirror, a first collimating mirror, a second a dispersive element and a second imaging mirror; the first encoding template is used to perform spectral dimension encoding on spectral image information of different fields of view imaged by the second imaging mirror; the light combining unit includes an exit light path along the first encoding template a second collimating mirror, a second dispersive element, and a third imaging mirror arranged in sequence in directions; wherein, the second dispersive element is used to combine the light beams dispersed by the first dispersive element; the first dispersive element and the second dispersive element The elements have the same structure and are symmetrically distributed along the first encoding template; the second encoding template is used to perform spatial micro-encoding on the spatial image information imaged by the third imaging mirror; the imaging mirror group is used to encode the encoded image Compressed to the focal plane of the unit detector; the unit detector is used to receive and collect the encoded and compressed information, and the data processing unit is used to invert the information obtained by the unit detector to realize the spatial information and spectrum of the target. Restoration of information, obtaining the 3D data cube of the target.

Further, the imaging lens group includes two mutually perpendicular cylindrical mirrors arranged in sequence along the beam direction, which are respectively a first cylindrical mirror and a second cylindrical mirror, and the first cylindrical mirror compresses the encoded image. into a line, and the second cylindrical mirror compresses the line into point information;

Or the imaging lens group includes one or more lenses.

Further, the apertures and focal lengths of the first dispersion element and the second dispersion element are the same.

Further, the first dispersion element and the second dispersion element are both prisms or gratings.

Further, both the first coding template and the second coding template include a movable mechanical template and a fixed coding template; the fixed coding template includes a digital micromirror array and a liquid crystal spatial light modulator; the first coding template and the first coding template The coding system of the two-coded template is Hadamard coding, Fourier transform coding or random coding.

At the same time, the present invention provides a gaze-type spectral imaging system based on compressed sensing, which is special in that it includes a first imaging mirror, a second encoding template, a spectroscopic unit, a first encoding template, a a light unit, an imaging mirror group, a unit detector and a data processing unit; the first imaging mirror is used to image the target at the primary image plane position; the second encoding template is used to image the space image after the first imaging mirror The information is spatially micro-encoded; the light splitting unit includes a first collimating mirror, a first dispersion element, and a second imaging mirror arranged in sequence along the direction of the exit light path of the second encoding template; the first encoding template is used for the second encoding template. Spectral image information of different fields of view imaged by the imaging mirror is subjected to spectral dimension encoding; the light combining unit includes a second collimating mirror, a second dispersive element, and a third imaging mirror sequentially arranged along the direction of the exit light path of the first encoding template; The second dispersive element is used to combine the beams dispersed by the first dispersive element; the first and second dispersive elements have the same structure and are symmetrically distributed along the first encoding template; the imaging lens group uses is used to compress the image imaged by the third imaging mirror onto the focal plane of the unit detector; the unit detector is used for receiving and collecting the compressed information, and the data processing unit is used for processing the information acquired by the unit detector. Inversion, realizes the restoration of the spatial information and spectral information of the target, and obtains the three-dimensional data cube of the target.

Further, the imaging lens group includes two mutually perpendicular cylindrical mirrors arranged in sequence along the beam direction, namely a first cylindrical mirror and a second cylindrical mirror, and the first cylindrical mirror compresses the image after imaging. into a line, and the second cylindrical mirror compresses the line into point information;

Or the imaging lens group includes one or more lenses.

Further, the first dispersion element and the second dispersion element are both prisms or gratings.

Further, the apertures and focal lengths of the first dispersion element and the second dispersion element are the same.

Further, both the first coding template and the second coding template include a movable mechanical template and a fixed coding template; the fixed coding template includes a digital micromirror array and a liquid crystal spatial light modulator; the first coding template and the first coding template The coding system of the two-coded template is Hadamard coding, Fourier transform coding or random coding.

Compared with the prior art, the advantages of the present invention are:

1. The spectral imaging system of the present invention firstly encodes the target spectral dimension information to obtain a spectrally encoded compressed image, and then encodes the target spatial dimension information, or first encodes the target spatial dimension information to obtain a spatially encoded compressed image , and then encode the spectral dimension information of the target, and finally obtain the compressed data of the entire spectrum by the unit detector, complete the encoding of the spectrum and space, so as to realize single-pixel spectral imaging; the compressed data contains both the image information and spectrum of the target. information, and the solution can realize gaze imaging, and the spatial information of the target can be obtained without scanning. The imaging system is small in size, low in cost and good in imaging quality;

The process of encoding the target spectral dimension can realize that the target spectral information obtained by one imaging has multiple spectral channel information, and at the same time, it can ensure that the obtained target image information has a high signal-to-noise ratio;

The encoding and compression process can compress the data, greatly reducing the data storage, acquisition and transmission time.

2. The spectral imaging system of the present invention can be applied to visible light, infrared, terahertz, microwave and other wavelength bands, and has wide applicability.

3. In the present invention, a data processing unit can be used to invert the information acquired by the unit detector to acquire a three-dimensional data cube of the target.

Description of drawings

Fig. 1 is the optical path diagram of the first embodiment of the staring-type spectral imaging system based on compressed sensing of the present invention;

FIG. 2 is an optical path diagram of the second embodiment of the gaze-type spectral imaging system based on compressed sensing according to the present invention.

Among them, the reference numerals are as follows:

1-first imaging mirror, 2-beam splitting unit, 21-first collimating mirror, 22-first dispersive element, 23-second imaging mirror, 3-first encoding template, 4-light combining unit, 41-th Two collimating mirrors, 42-second dispersive element, 43-third imaging mirror, 5-second encoding template, 6-imaging mirror group, 7-unit detector.

Detailed ways

The content of the present invention will be described in further detail below with reference to the accompanying drawings and specific embodiments.

Example 1

The spectral imaging technology proposed in this embodiment first realizes the spectral dimension encoding of the target through the spatial light modulator. On this basis, the spatial dimension encoding of the target is performed again, and finally the encoded information is sampled through a single pixel to achieve a single pixel. Gaze spectral imaging. The spatial image information and spectral information of the target can be recovered by inverting the collected information through the compressed sensing algorithm. Compared with large-scale area array detectors, unit detectors often have higher performance, including lower detector noise, higher quantum efficiency and detection frame rate, and larger detection signal receiving area. At the same time, the three-dimensional data of the atlas of the target can be obtained through the unit detector.

As shown in FIG. 1, a gaze-type spectral imaging system based on compressed sensing includes a first imaging mirror 1, a light splitting unit 2, a first encoding template 3, a light combining unit 4, and a second encoding template arranged in sequence along the beam direction 5. Imaging lens group 6, unit detector 7 and data processing unit.

The first imaging mirror 1 is composed of a single piece or a plurality of mirrors, and is used to image the target on the primary image plane position.

The beam splitting unit 2 includes a first collimating mirror 21, a first dispersive element 22, and a second imaging mirror 23 arranged in sequence along the direction of the exit light path of the first imaging mirror 1; wherein, the first collimating mirror 21 is composed of a single piece or a plurality of It is composed of lenses, which collimate the light beams entering the system, so that the incident light beams in different fields of view enter the first dispersive element 22 in parallel; , the dispersed light beam propagates to the second imaging mirror 23, and the dispersed light beam through the first dispersive element 22 contains the spectral information of the target; the second imaging mirror 23 is composed of a single piece or multiple mirrors to image the dispersed light beam, Spectral information of the target is obtained at the first encoding template 3 .

The first encoding template 3 is used to perform spectral dimension encoding on the spectral information of different fields of view of the target, realize the spectral dimension encoding of the target, and obtain an image encoded by the spectral dimension; the encoding template includes a spatial light modulation device digital micromirror array (DMD), Liquid crystal spatial light modulator (LCM) and mechanical mask, wherein the coding system includes Hadamard coding, Fourier transform coding and random coding and other coding forms.

The light combining unit 4 includes a second collimating mirror 41, a second dispersive element 42, and a third imaging mirror 43 that are sequentially arranged along the direction of the exit light path of the first encoding template; the second collimating mirror 41 is composed of a single piece or multiple mirrors The coded beams of different spectral bands are collimated, and then enter the second dispersive element 42; the function of the second dispersive element 42 is to combine the beams of different wavelength bands dispersed by the first dispersive element 22. The light beam propagates to the third imaging mirror 43; wherein the second dispersive element 42 and the first dispersive element 22 have the same model and parameters, which can be gratings or prisms; the third imaging mirror 43 images the encoded light beam, and the second The spatial image of the target is obtained at the position of coding template 5. The first dispersion element 22 and the second dispersion element 42 have the same structure and are symmetrically distributed along the first encoding template 3 .

The second coding template 5 is a spatial coding template, which is used to perform spatial microcoding on the spatial image encoded by the spectral dimension; it includes various forms such as a digital micromirror array (DMD), a liquid crystal spatial light modulator (LCM), and a mechanical template. ; The coding system includes Hadamard coding, Fourier transform coding and random coding and other coding forms.

The imaging lens group 6 is used to compress the encoded image to the focal plane of the unit detector 7. The imaging lens group 6 has a special function to zoom the entire image to a very small scale; the imaging lens group 6 is composed of a single piece. Or composed of multiple lenses, you can use two mutually perpendicular cylindrical mirrors arranged in sequence along the beam direction, compress the encoded image into a line through the first cylindrical mirror, and then use the second cylindrical mirror to compress the line. Compressed into point information; or the imaging lens group 6 uses a lens lens group to compress the image.

The unit detector 7 is used to receive and collect the encoded and compressed information.

Data processing unit, signal recovery algorithm based on compressed sensing, such as matching pursuit (MP), orthogonal matching pursuit (OMP), threshold hard iteration (IHT), compressed sampling matching pursuit (CoSaMP) and subspace pursuit (SP), through The compressed sensing algorithm inverts the information obtained by the unit detector 7, which can restore the spatial information and spectral information of the target, and obtain the three-dimensional data cube of the target.

The spectral dimension information of the target is encoded by the coding imaging technology, and the compressed image of the spectral dimension encoding can be obtained. The dispersion can use various light-splitting elements such as prisms and gratings, and the encoding can use various Hadamard coding, Fourier coding and their derivatives Encoding form; on this basis, the spatial dimension information is encoded through the encoding template, and finally the compressed data of the entire atlas is obtained from the pixel structure of the unit detector. The compressed data contains both the image information and spectral information of the target, and the solution can realize gaze imaging, and the spatial information of the target can be obtained without scanning.

The spectral imaging system of this embodiment realizes the encoding of spectrum and space through the technical solution of double encoding, thereby realizing single-pixel spectral imaging.

This embodiment realizes the encoding of the spectral information of the target through the dual-dispersion aperture encoding technology (including two dispersive elements), during which the complete image information of the target is obtained at the same time, and on this basis, the obtained image information is encoded, After compression, it is received by the unit detector to realize staring single-pixel spectral imaging technology.

Compared with the existing single-pixel spectral imaging technology, the spectral imaging system of this embodiment has a higher signal-to-noise ratio. The spectral dimension coding is realized by the dual-dispersion aperture coding spectral imaging technology. The target spectral information obtained by one imaging in this process has multiple spectral channel information. It is a multi-channel spectral imaging technology, which can ensure that the obtained target image information has high signal-to-noise ratio.

The spectral imaging system of this embodiment is suitable for multiple wavelength bands including visible light, infrared, terahertz, microwave, etc. Compared with other single-pixel spectral imaging devices, the coding template of this technology can use various spatial light modulation devices, including digital micromirror devices. (DMD), liquid crystal spatial light modulator, mechanical mask and other devices, with wide applicability.

The spectral imaging system of this embodiment has a great data compression function. Compared with the existing technical solutions, this technology can achieve a data compression ratio dozens of times higher than the existing solutions, greatly reducing the time and difficulty of data storage, collection and transmission.

Embodiment 2

The spectral imaging technology proposed in this embodiment first realizes the spatial dimensional coding of the target through the spatial light modulator, and then performs the spectral dimensional coding of the target again on this basis, and finally samples the coded information through a single pixel to realize a single pixel. Gaze spectral imaging. The spatial image information and spectral information of the target can be recovered by inverting the collected information through the compressed sensing algorithm. Compared with large-scale area array detectors, unit detectors often have higher performance, including lower detector noise, higher quantum efficiency and detection frame rate, and larger detection signal receiving area. At the same time, the three-dimensional data of the atlas of the target can be obtained through the unit detector.

As shown in FIG. 2, a gaze-type spectral imaging system based on compressed sensing includes a first imaging mirror 1, a second encoding template 5, a light splitting unit 2, a first encoding template 3, and a light combining unit arranged in sequence along the beam direction 4. Imaging mirror group 6 and unit detector 7 .

The first imaging mirror 1 is composed of a single piece or a plurality of mirrors, and is used to image the target on the primary image plane position.

The second encoding template 5 is a spatial encoding template, which is used to perform spatial micro-encoding on the spatial image information imaged by the first imaging mirror 1; it includes a digital micromirror array (DMD), a liquid crystal spatial light modulator (LCM) and a mechanical template Among them, the coding system includes various coding forms such as Hadamard coding, Fourier transform coding and random coding.

The beam splitting unit 2 includes a first collimating mirror 21, a first dispersive element 22, and a second imaging mirror 23 arranged in sequence along the direction of the exit light path of the second encoding template 5; wherein, the first collimating mirror 21 is composed of a single piece or a plurality of It is composed of lenses, which collimate the light beams entering the system, so that the incident light beams in different fields of view enter the first dispersive element 22 in parallel; , the dispersed light beam propagates to the second imaging mirror 23, and the dispersed light beam through the first dispersive element 22 contains the spectral information of the target; the second imaging mirror 23 is composed of a single piece or multiple mirrors to image the dispersed light beam, Spectral information of the target is obtained at the first encoding template 3 .

The first encoding template 3 is used to perform spectral dimension encoding on the spectral information of different fields of view of the target, realize the spectral dimension encoding of the target, and obtain a spectrally encoded image; the encoding template includes a spatial light modulation device digital micromirror array (DMD) and Liquid crystal spatial light modulator (LCM) and mechanical mask, wherein the coding system includes Hadamard coding, Fourier transform coding and random coding and other coding forms.

The light combining unit 4 includes a second collimating mirror 41, a second dispersive element 42, and a third imaging mirror 43 that are sequentially arranged along the direction of the exit light path of the first encoding template; the second collimating mirror 41 is composed of a single piece or multiple mirrors The coded beams of different spectral bands are collimated, and then enter the second dispersive element 42; the function of the second dispersive element 42 is to combine the beams of different wavelength bands dispersed by the first dispersive element 22. The light beam propagates to the third imaging mirror 43; wherein the second dispersive element 42 and the first dispersive element 22 have the same model and parameters, which can be gratings or prisms; the third imaging mirror 43 images the encoded light beam, and the second The spatial image of the target is obtained at the position of coding template 5. The first dispersion element 22 and the second dispersion element 42 have the same structure and are symmetrically distributed along the first encoding template 3 .

The imaging mirror group 6 is used to compress the coded image to the focal plane of the unit detector 7. The imaging mirror group 6 has a special function to zoom the entire image to a small scale; the imaging mirror group 6 is composed of a single unit. It can be composed of two or more lenses. Two mutually perpendicular cylindrical mirrors can be used in sequence along the beam direction. The first cylindrical mirror compresses the encoded image into a line, and then the second cylindrical mirror compresses the coded image into a line. Lines are compressed into point information; or the imaging lens group 6 uses a lens lens group to compress the image.

The unit detector 7 is used to receive and collect the encoded and compressed information.

Data processing unit, data processing unit based on compressed sensing (signal recovery algorithm), such as matching pursuit (MP), orthogonal matching pursuit (OMP), threshold hard iteration (IHT), compressed sampling matching pursuit (CoSaMP) and subspace tracking (SP), by inverting the information obtained by the unit detector 7 through the compressed sensing algorithm, the spatial information and spectral information of the target can be restored, and the three-dimensional data cube of the target can be obtained.

The target spatial dimension information is encoded by the coding imaging technology, and the compressed image of the spatial dimension coding can be obtained, in which the dispersion can use various light-splitting elements such as prisms and gratings, and the coding can use various kinds of Hadamard coding, Fourier coding and their derivatives. Encoding form; on this basis, the spectral dimension information is encoded through the encoding template, and finally the compressed data of the entire spectrum is obtained from the pixel structure of the unit detector. The compressed data contains both the image information and spectral information of the target, and the solution can realize gaze imaging, and the spatial information of the target can be obtained without scanning.

The spectral imaging system of this embodiment realizes the encoding of spectrum and space through the technical solution of double encoding, thereby realizing single-pixel spectral imaging.

This embodiment realizes the encoding of the spectral information of the target through the dual-dispersion aperture encoding technology (including two dispersive elements), during which the complete image information of the target is obtained at the same time, and on this basis, the obtained image information is encoded, After compression, it is received by the unit detector to realize staring single-pixel spectral imaging technology.

Compared with the existing single-pixel spectral imaging technology, the spectral imaging system of this embodiment has a higher signal-to-noise ratio. The spectral dimension coding is realized by the dual-dispersion aperture coding spectral imaging technology. The target spectral information obtained by one imaging in this process has multiple spectral channel information. It is a multi-channel spectral imaging technology, which can ensure that the obtained target image information has high signal-to-noise ratio.

The spectral imaging system of this embodiment is suitable for multiple wavelength bands including visible light, infrared, terahertz, microwave, etc. Compared with other single-pixel spectral imaging devices, the coding template of this technology can use various spatial light modulation devices, including digital micromirror devices. (DMD), liquid crystal spatial light modulator, mechanical mask and other devices, with wide applicability.

The spectral imaging system of this embodiment has a great data compression function. Compared with the existing technical solutions, this technology can achieve a data compression ratio dozens of times higher than the existing solutions, greatly reducing the time and difficulty of data storage, collection and transmission.

The above only describes the preferred embodiments of the present invention, and does not limit the technical solutions of the present invention to this. Any known deformations made by those skilled in the art on the basis of the main technical concept of the present invention belong to the protection of the present invention. technical category.

Claims (10)

1. A gaze-type spectral imaging system based on compressed sensing, characterized in that: it comprises a first imaging mirror (1), a light-splitting unit (2), a first encoding template (3), a light-combining unit that are arranged in turn along the beam direction (4), a second encoding template (5), an imaging lens group (6), a unit detector (7) and a data processing unit; The first imaging mirror (1) is used to image the target at the primary image plane position; The spectroscopic unit (2) comprises a first collimating mirror (21), a first dispersive element (22), and a second imaging mirror (23) that are sequentially arranged along the direction of the outgoing light path of the first imaging mirror (1); The first encoding template (3) is used to perform spectral dimension encoding on spectral image information of different fields of view imaged by the second imaging mirror (23); The light combining unit (4) comprises a second collimating mirror (41), a second dispersive element (42), and a third imaging mirror (43) arranged in sequence along the direction of the outgoing light path of the first encoding template (3); wherein , the second dispersion element (42) is used to combine the light beams dispersed by the first dispersion element (22); The first dispersion element (22) and the second dispersion element (42) have the same structure and are symmetrically distributed along the first encoding template (3); The second encoding template (5) is used to perform spatial micro-encoding on the spatial image information imaged by the third imaging mirror (43); The imaging lens group (6) is used for compressing the encoded image onto the focal plane of the unit detector (7); The unit detector (7) is used for receiving and collecting the encoded and compressed information; The data processing unit is used for inverting the information obtained by the unit detector (7), realizing the restoration of the spatial information and spectral information of the target, and obtaining the three-dimensional data cube of the target. 2. The staring-type spectral imaging system based on compressed sensing according to claim 1, wherein the imaging mirror group (6) comprises two mutually perpendicular cylindrical mirrors arranged in sequence along the beam direction, which are the first A cylindrical mirror and a second cylindrical mirror, the first cylindrical mirror compresses the encoded image into a line, and the second cylindrical mirror compresses the line into point information; Or the imaging lens group (6) includes one or more lenses. 3. The staring-type spectral imaging system based on compressed sensing according to claim 1, characterized in that: the first dispersive element (22) and the second dispersive element (42) have equal apertures and equal focal lengths. 4. The gaze-type spectral imaging system based on compressed sensing according to claim 1, wherein the first dispersion element (22) and the second dispersion element (42) are both prisms or gratings. 5. The gaze-type spectral imaging system based on compressed sensing according to any one of claims 1 to 4, wherein the first encoding template (3) and the second encoding template (5) both comprise a mobile mechanical template and Fixed coding templates; The fixed coding template includes a digital micromirror array and a liquid crystal spatial light modulator; The coding systems of the first coding template (3) and the second coding template (5) are Hadamard coding, Fourier transform coding or random coding. 6. A gaze-type spectral imaging system based on compressive sensing, characterized in that it comprises a first imaging mirror (1), a second encoding template (5), a spectroscopic unit (2), a first encoding that are arranged in sequence along the beam direction a template (3), a light combining unit (4), an imaging lens group (6), a unit detector (7) and a data processing unit; The first imaging mirror (1) is used to image the target at the primary image plane position; The second encoding template (5) is used to perform spatial micro-encoding on the spatial image information imaged by the first imaging mirror (1); The spectroscopic unit (2) comprises a first collimating mirror (21), a first dispersion element (22), and a second imaging mirror (23) arranged in sequence along the direction of the exit light path of the second encoding template (5); The first encoding template (3) is used to perform spectral dimension encoding on spectral image information of different fields of view imaged by the second imaging mirror (23); The light combining unit (4) comprises a second collimating mirror (41), a second dispersive element (42), and a third imaging mirror (43) arranged in sequence along the direction of the outgoing light path of the first encoding template (3); wherein , the second dispersion element (42) is used to combine the light beams dispersed by the first dispersion element (22); The first dispersion element (22) and the second dispersion element (42) have the same structure and are symmetrically distributed along the first encoding template (3); The imaging mirror group (6) is used for compressing the image formed by the third imaging mirror (43) onto the focal plane of the unit detector (7); The unit detector (7) is used for receiving and collecting the compressed information; The data processing unit is used for inverting the information obtained by the unit detector (7), realizing the restoration of the spatial information and spectral information of the target, and obtaining the three-dimensional data cube of the target. 7. The staring-type spectral imaging system based on compressed sensing according to claim 6, wherein the imaging mirror group (6) comprises two mutually perpendicular cylindrical mirrors arranged in sequence along the beam direction, which are the first A cylindrical mirror and a second cylindrical mirror, the first cylindrical mirror compresses the image after imaging into a line, and the second cylindrical mirror compresses the line into point information; Or the imaging lens group (6) includes one or more lenses. 8. The gaze-type spectral imaging system based on compressed sensing according to claim 6, wherein the first dispersion element (22) and the second dispersion element (42) are both prisms or gratings. 9. The gaze-type spectral imaging system based on compressed sensing according to claim 6, characterized in that: the first dispersive element (22) and the second dispersive element (42) have equal apertures and equal focal lengths. 10. The gaze-type spectral imaging system based on compressed sensing according to any one of claims 6 to 9, wherein: The first encoding template (3) and the second encoding template (5) both include a mobile mechanical template and a fixed encoding template; The fixed coding template includes a digital micromirror array and a liquid crystal spatial light modulator; The coding systems of the first coding template (3) and the second coding template (5) are Hadamard coding, Fourier transform coding or random coding.

Images